The microbiota of the human vagina can profoundly affect the health of women and neonates. For instance, women with the condition bacterial vaginosis (BV) have increased risks of acquiring sexually transmitted infections such as HIV, developing pelvic inflammatory disease, and experiencing preterm birth. Healthcare costs associated with BV are enormous, whether considering the direct costs related to the millions of healthcare visits annually for vaginal discharge, or the indirect costs such as care of pre-term infants. Despite recent advances that have used molecular methods (PCR) to better describe the microbiology of BV, there is a lack of understanding regarding why some women develop this condition. Changes in the vaginal microbiota after women develop BV are well described, but the microbiological, physiological, and behavioral antecedents to BV are poorly understood and connected. This proposal seeks to overcome this knowledge gap by collecting data on women at risk for BV in a longitudinal study designed to capture incident and relapsing cases. Broad range 16S rRNA gene PCR with pyrosequencing and bacterium-specific quantitative PCR will be used to characterize vaginal microbiotas. Changes in vaginal bacterial communities will be linked to physiological factors (menses), medication use (hormonal contraceptives, antibiotics), and behavioral factors (sexual practices, condom use). Studying the factors that lead to BV is critical for understanding pathogenesis. This knowledge can be used to develop improved methods for diagnosis, prevention, and treatment. Although antibiotics such as metronidazole and clindamycin can be used to treat BV, recurrence is common, with recurrence rates of up to 30% at 3 months and >50% at one year. The reasons for this high rate of recurrence are not known. High relapse rates complicate efforts to ameliorate the adverse health consequences of BV. This proposal seeks to describe the microbiological factors associated with recurrence of BV by studying extra-vaginal reservoirs of bacteria that may allow seeding of the vagina. Women with BV (and a subset of sex partners) will be followed during and after antibiotic treatment to assess concentrations of vaginal bacteria in vaginal and extra-vaginal sites using PCR. It is anticipated that colonization of extra-vaginal reservoirs will be a major risk factor for recurrence of BV. There is ongoing debate whether biofilms are important in the pathogenesis of BV, with intriguing data supporting this concept based on some small cross-sectional studies. Fluorescence in situ hybridization (FISH) will be used to assess the prevalence of bacterial biofilms in vaginal samples obtained from women with and without BV. The value of non-biopsy samples will be studied here. Changes in vaginal bacterial biofilms will be assessed after antibiotic treatment to determine if recurrence of BV is associated with persistence of bacteria in biofilm communities. If biofilms are shown to be important in BV, this would open many new avenues for treatment and prevention based on disruption of biofilm formation and cell signaling.